Javier Chávarri

I see, thanks for the pointers. I could find the `test/blackbox-tests/test-cases/merlin/default-based-context.t` test, which contains similar checks than the ones added in this PR. For additional context 😅 I am investigating...

Thanks for the pointers, will take a look. > likely it will not be enough because `enabled_if` on a library accepts very few arguments. That's a different limitation that I...

Sharing here some notes of what I've been able to find out. Currently, there is a uniform "view" (for lack of a better name) of library dbs that is shared...

I'm leaving here some notes, as I have very limited idea of what i'm doing, so maybe they can serve to surface if I'm going in the wrong direction: -...

@rgrinberg Thanks for the pointers, they are super helpful. I made some progress on the new handling of lib dbs following the steps mentioned (see https://github.com/ocaml/dune/pull/9839/commits/1b0c2c33f9c92cf51c7d345d6112927f0fb4f870). Also fixed a cryptic...

I have been looking into a weird issue. When I defined a test using a public library using the code in this PR: ``` $ cat > dune-project (lang dune...

Unrelated to the comment above (sorry for the unordered stream of comments). I have some suspicions that this calculation of `enabled_if` is not correct: https://github.com/ocaml/dune/blob/21557d8d0b68650d4997c5ec94a1a782a2847a87/src/dune_rules/gen_rules.ml#L117 The reason is that it...

In https://github.com/ocaml/dune/pull/9839/commits/548fabd70d13d9e68d6fc9c851335e42dcda6694, I updated the test to show the issue I've been trying to figure out for the last few days.

@rgrinberg I hit another issue with the current approach we're taking. The problem is that now we are checking for duplicated libraries at DB read time (`Lib.DB.find`) rather than DB...

@rgrinberg I think I got the PR to a point where the feature and error handling seems to be working, as far as I can tell. Regarding my last comment...